Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import PyPDF2
|
| 3 |
+
|
| 4 |
+
def extract_pdf_lines(pdf_file):
|
| 5 |
+
# Extract all lines with page/line numbers for user reference
|
| 6 |
+
with open(pdf_file.name, "rb") as f:
|
| 7 |
+
reader = PyPDF2.PdfReader(f)
|
| 8 |
+
result = []
|
| 9 |
+
for i, page in enumerate(reader.pages):
|
| 10 |
+
page_text = page.extract_text()
|
| 11 |
+
if page_text:
|
| 12 |
+
lines = page_text.splitlines()
|
| 13 |
+
for ln, line in enumerate(lines):
|
| 14 |
+
# Prefix each with [Page X Line Y] for clarity
|
| 15 |
+
result.append(f"[Page {i+1} Line {ln+1}] {line}")
|
| 16 |
+
return "\n".join(result) if result else "[NO TEXT FOUND]"
|
| 17 |
+
|
| 18 |
+
def show_sample_context(raw_text, example):
|
| 19 |
+
"""Show where sample value occurs for user feedback and future rule logic."""
|
| 20 |
+
context_lines = []
|
| 21 |
+
lines = raw_text.splitlines()
|
| 22 |
+
example = example.strip()
|
| 23 |
+
for i, line in enumerate(lines):
|
| 24 |
+
if example and example in line:
|
| 25 |
+
prev_line = lines[i-1] if i > 0 else ""
|
| 26 |
+
next_line = lines[i+1] if i+1 < len(lines) else ""
|
| 27 |
+
snippet = f"...\n{prev_line}\n>>> {line} <<<\n{next_line}\n..."
|
| 28 |
+
context_lines.append(snippet)
|
| 29 |
+
if not context_lines:
|
| 30 |
+
return "No match for example in extracted text."
|
| 31 |
+
return "\n---\n".join(context_lines)
|
| 32 |
+
|
| 33 |
+
with gr.Blocks() as demo:
|
| 34 |
+
gr.Markdown("# PDF Field Teach-Me Extractor (Hugging Face Space Example)")
|
| 35 |
+
pdf_in = gr.File(label="Upload PDF")
|
| 36 |
+
raw_text = gr.Textbox(
|
| 37 |
+
label="Raw extracted text (copy-paste any value for training)",
|
| 38 |
+
lines=20,
|
| 39 |
+
interactive=False,
|
| 40 |
+
show_copy_button=True
|
| 41 |
+
)
|
| 42 |
+
pdf_in.change(extract_pdf_lines, inputs=pdf_in, outputs=raw_text)
|
| 43 |
+
|
| 44 |
+
gr.Markdown("""
|
| 45 |
+
### 2️⃣ Teach the System
|
| 46 |
+
- Copy a sample value you want to extract, and paste it below.
|
| 47 |
+
- Assign a field name (will become a column header later).
|
| 48 |
+
- Click "Show context" to see where this value appears (future step: use this to build automatic rules!).
|
| 49 |
+
""")
|
| 50 |
+
label = gr.Textbox(label="Field Name (e.g., 'Customer Name')", lines=1)
|
| 51 |
+
example = gr.Textbox(label="Sample Value (copy from above)", lines=1)
|
| 52 |
+
context_out = gr.Textbox(label="Value context in raw PDF text", lines=4)
|
| 53 |
+
search_btn = gr.Button("Show context")
|
| 54 |
+
|
| 55 |
+
search_btn.click(fn=show_sample_context, inputs=[raw_text, example], outputs=context_out)
|
| 56 |
+
|
| 57 |
+
demo.launch()
|